ORIGINAL
ARTICLE
Effect of Vitamin D3 Supplementation in Black and in White Children: A Randomized, Placebo-Controlled Trial Kumaravel Rajakumar, Charity G. Moore, Jonathan Yabes, Flora Olabopo, Mary Ann Haralam, Diane Comer, Jaimee Bogusz, Anita Nucci, Susan Sereika, Jacqueline Dunbar-Jacob, Michael F. Holick, and Susan L. Greenspan Department of Pediatrics (K.R., F.O., M.A.H.), and Center for Research on Health Care (C.G.M., J.Y., D.C.), University of Pittsburgh, Pittsburgh, Pennsylvania 15213; Department of Medicine (J.B., M.F.H.), Boston University School of Medicine, Boston, Massachusetts, 02118; Department of Nutrition (A.N.), Georgia State University, Atlanta, Georgia 30302; University of Pittsburgh School of Nursing (S.S., J.D.-J.), Pittsburgh, Pennsylvania 15213; and Department of Medicine (C.G.M., J.Y., D.C., S.L.G.), University of Pittsburgh, Pittsburgh 15213
Context: Dosages of vitamin D necessary to prevent or treat vitamin D deficiency in children remain to be clarified. Objective: To determine the effects of vitamin D3 1000 IU/d on serum 25-hydroxyvitamin D [25(OH)D], PTH, and markers of bone turnover (osteocalcin and collagen type 1 cross-linked Ctelopeptide) in black children and white children, and to explore whether there is a threshold level of 25(OH)D associated with maximal suppression of serum PTH concentration. Design: Healthy 8- to 14-year-old Pittsburgh-area black (n ⫽ 84) and white (n ⫽ 73) children not receiving vitamin supplements, enrolled from October through March from 2008 through 2011, were randomized to vitamin D3 1000 IU or placebo daily for 6 months. Results: The mean baseline concentration of 25(OH)D was ⬍20 ng/mL in both the vitamin D-supplemented group and the placebo group (19.8 ⫾ 7.6 and 18.8 ⫾ 6.9 ng/mL, respectively). The mean concentration was higher in the supplemented group than in the placebo group at 2 months (26.4 ⫾ 8.1 vs 18.9 ⫾ 8.1 ng/mL; P ⬍ .0001) and also at 6 months (26.7 ⫾ 7.6 vs 22.4 ⫾ 7.3; P ⫽ .003), after adjusting for baseline 25(OH)D, race, gender, pubertal status, dietary vitamin D intake, body mass index, and sunlight exposure. Increases were only significant in black children, when examined by race. The association between 25(OH)D and PTH concentrations was inverse and linear, without evidence of a plateau. Overall, vitamin D supplementation had no effect on PTH and bone turnover. Conclusions: Vitamin D3 supplementation with 1000 IU/d in children with mean baseline 25(OH)D concentration ⬍20 ng/mL effectively raised their mean 25(OH)D concentration to ⱖ20 ng/mL but failed to reach 30 ng/mL. Vitamin D supplementation had no effect on PTH concentrations. (J Clin Endocrinol Metab 100: 3183–3192, 2015)
aintaining adequate vitamin D status is essential for calcium homeostasis and skeletal health. However, hypovitaminosis D is common in healthy children living in the northeastern United States, and its prevalence
M
and severity are greater in black than in white children (1). Circulating concentration of 25-hydroxyvitamin D [25(OH)D] is the recognized biomarker of vitamin D status. Definition of a 25(OH)D cutoff level for optimal skel-
ISSN Print 0021-972X ISSN Online 1945-7197 Printed in USA Copyright © 2015 by the Endocrine Society Received March 11, 2015. Accepted June 16, 2015. First Published Online June 19, 2015
Abbreviations: BMI, body mass index; CTx, collagen type 1 cross-linked C-telopeptide; CV, coefficient of variation; DBP, vitamin D-binding-protein; OC, osteocalcin; 25(OH)D, 25hydroxyvitamin D; ⌬, change.
doi: 10.1210/jc.2015-1643
J Clin Endocrinol Metab, August 2015, 100(8):3183–3192
press.endocrine.org/journal/jcem
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 August 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.
3183
3184
Rajakumar et al
Vitamin D Supplementation in Children
etal health lacks consensus. The Institute of Medicine (IOM) recommends concentrations of 25(OH)D ⱖ20 ng/mL as optimal for skeletal health and defines vitamin D deficiency as a concentration ⬍12 ng/mL and vitamin D insufficiency as concentrations 12 to 20 ng/mL (2). The IOM estimates that concentrations of ⱖ16 ng/mL and ⱖ20 ng/mL would be adequate for ensuring the skeletal health needs of 50 and 97.5%, respectively, of US children. The Endocrine Society and other experts in the field have suggested a target level ⱖ30 ng/mL for optimal skeletal health (3, 4). Concentrations of 25(OH)D associated with maximal suppression of serum PTH concentration, indicated by a plateauing of the PTH concentration, have been used for defining vitamin D sufficiency in adults. Such data in children are limited and varied. Hill et al (5) found no inflection point in the inverse association between 25(OH)D and PTH in a cross-sectional study of 7- to 18-year-old children. Although Maguire et al (6) demonstrated plateauing of PTH at a 25(OH)D concentration of 42.8 ng/mL in a cross-sectional study of 1- to 6-year-old children, vitamin D supplementation (0, 400, 1000, 2000, and 4000 IU/d for 12 wk) in 9- to 13-year-old children had no effect on PTH concentrations despite dose-dependent increases in 25(OH)D concentrations (7). Confounding effects of sun exposure and other determinants of vitamin D photoproduction pose challenges for estimating the amount of dietary vitamin D needed to achieve and maintain a defined 25(OH)D level (2). Those considerations notwithstanding, the IOM’s dietary reference intakes for vitamin D were calculated without regard for variations in skin color, race, or sunlight exposure. Delineating the racial differences in response to vitamin D supplementation and exploring levels of 25(OH)D associated with vitamin D sufficiency is relevant for formulating dietary guidelines. In addition, data regarding the effect of vitamin D supplementation on bone turnover are limited. Therefore, we initiated a pharmacological challenge with 1000 IU of vitamin D3 daily (five times the prevailing adequate intake for vitamin D) (8) in black children and white children to examine the responsive changes in 25(OH)D, PTH, and bone turnover. The objective of this study was to determine the effects of supplementation with 1000 IU of vitamin D3 on serum concentrations of 25(OH)D, PTH, and markers of bone turnover in black children and white children. We utilized the longitudinal design of our study to determine whether there was a threshold level of 25(OH)D associated with maximal suppression of serum PTH concentration.
J Clin Endocrinol Metab, August 2015, 100(8):3183–3192
Subjects and Methods Study design and participants We enrolled healthy 8- to 14-year-old children from Pittsburgh and Kittanning, Pennsylvania (latitude/longitude: 40.4° N/80°W and 40.8° N/79.5°W, respectively) in a randomized, double-blind, placebo-controlled trial of vitamin D3 1000 IU daily (Clinicaltrials.gov identifier: NCT00732758). The children were enrolled October through March of 2008 through 2011. Children receiving vitamin preparations underwent a 1-month washout before enrollment. Subjects were recruited from the Primary Care Center of the Children’s Hospital of Pittsburgh of the University of Pittsburgh Medical Center (UPMC) and the Children’s Community Pediatrics-Armstrong practice in Kittanning, and through advertisements posted in the offices of the Children’s Hospital-affiliated Pediatric PittNet, a practice-based pediatric research network. Study procedures subsequent to enrollment were conducted either at the UPMC Montefiore Clinical and Translational Research Center or at the Children’s Community Pediatrics-Armstrong practice. The study was approved by the University of Pittsburgh Institutional Review Board. Signed informed parental consent and subjects’ assent were obtained before research participation. The race of subjects was specified by their parents.
Randomization and intervention Randomization was stratified by race using a 1:1 allocation ratio and a permuted block scheme with block size of 4. Children received either vitamin D3 1000 IU or placebo in a single tablet once daily. The allocation scheme was generated by a study statistician using R version 2.7.0. The vitamin D3 and placebo tablets were manufactured by Douglas Laboratories, were similar in color, and were dispensed in identical containers labeled either A or B. A sealed envelope system was used for the assignments. Children, parents, and the investigative team were blinded to the treatment assignment until the study was completed. Study medications were made in two batches. The average vitamin D3 content of the vitamin D3 tablet, measured as previously described (9), in the first batch around the midpoint of its shelf-life was approximately 1129 IU; in the second batch at the end of the trial, it was approximately 1140 IU.
Compliance Compliance was assessed by pill count at the 2- and 6-month follow-up visits and was validated in a subset of 90 subjects by an electronic medication event monitoring system (MEMs 6 Track Cap; AARDEX).
Study measurements Anthropometry, sun exposure, skin color, dietary intake, and pubertal status We measured height and weight and calculated body mass index (BMI) at study entry and at the 6-month follow-up visit. At study entry and exit, we assessed: summertime sunlight exposure characteristics; melanin index from the forehead, back of the hand, and inner arm using a handheld dermatospectrophotometer (DSM II Colormeter; Cortex Technology); and dietary intake of vitamin D and calcium using a validated Youth and Adolescent Food Frequency Questionnaire (10, 11). We estimated
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 August 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2015-1643
press.endocrine.org/journal/jcem
Tanner stage by physical examination (12, 13) and ascertained the parent-reported Fitzpatrick sunreactive skin type (14 –16) at study entry.
3185
Study Flow Diagram
Assessed for eligibility (n=355)
Ineligible (n=51)
Biochemical assessments We collected blood samples by venipuncture in a nonfasting state throughout the day. We measured serum calcium, phosphorus, albumin, 25(OH)D, PTH, osteocalcin (OC; marker of bone formation), and collagen type 1 cross-linked C-telopeptide (CTx; a marker of bone resorption) concentrations at baseline and at 2- and 6-month follow-up visits. Calcium, phosphorus, and albumin concentrations were measured at the UPMC Clinical Chemistry laboratory. Serum 25(OH)D, PTH, OC, and CTx concentrations were measured at the Vitamin D, Skin, and Bone Research Laboratory at Boston University Medical Center. Serum 25(OH)D concentrations were measured using liquid chromatography-tandem mass spectrometry (17). The intra-assay and interassay coefficients of variation (CVs) were 6% and 10%, respectively. The 25(OH)D assay is Center for Disease Control certified, and National Institutes of Standards and Technology standards were used for confirmation of the standard curves. Serum 3-epi-25(OH)D is excluded in the reported levels. Serum PTH concentrations were assayed using a Human Bioactive PTH 1– 84 Elisa kit (Immutopics, Inc); intra-assay and interassay CVs were 7 and 9%, respectively. Serum OC concentrations were measured by Micro Vue OC enzyme immunoassay kit (Quidel); intra-assay and interassay CVs were 5 and 10%, respectively. Serum CTx concentrations were measured by Serum CrossLaps ELISA kit (Immunodiagnostic Systems); intra-assay and interassay CVs were 1.8 –3% and 8.0 –10.9%, respectively.
Statistical analysis We based parameter estimates for sample-size analyses on our earlier study of changes in PTH concentration in response to vitamin D3 supplementation (18). Assuming that the concentration of 25(OH)D would reach ⬎20 ng/mL in the supplemented group and would remain ⱕ20 ng/mL in the placebo group and with two-sided ␣ ⫽ 0.05, we estimated that 160 children would be needed to detect effect sizes in serum 25(OH)D and PTH concentrations of approximately 0.5, with power of 81 to 88%. Assuming that 50% of the children were black and 50% white, we had 79% power to detect a correlation of at least ⫺0.3 between PTH and 25(OH)D concentrations within each race group (␣ ⫽ 0.05). We anticipated an attrition rate of 5% and planned to enroll a total of 168 children. We used intention-to-treat analyses when testing the effects of vitamin D supplementation relative to placebo. We tested for treatment group or racial differences in categorical measures using 2 or Fisher’s exact tests and in continuous measures using t tests. Concentrations of 25(OH)D, the primary study outcome, did not require transformation. We compared mean 25(OH)D concentrations using analysis of covariance, controlling for baseline 25(OH)D, race, gender, pubertal status, BMI, dietary intake of vitamin D, and sunlight exposure. To explore threshold effects of 25(OH)D on PTH concentrations, we conducted analyses similar to those described by Hill et al (5) and Maguire et al (6) using linear and cubic spline regressions. Scatterplots were constructed superimposed with both fitted lines and Lowess (locally weighted scatterplot smoothing) curves to visually examine nonlinearity. We used mixed-effects modeling to examine the associations between
Eligible (n=304)
Did not agree to parcipate (n=147)
Consent form signed (n=157)
Randomized (n=157)
Vitamin D Group (n=78) Black, 42; White, 36
Lost to follow-up (n=8)
Completed (n=70)
Placebo Group (n=79) Black, 42; White, 37
Lost to follow-up (n=16)
Completed (n=63)
Figure 1. Enrollment, randomization, and follow-up of study participants
PTH and 25(OH)D concentrations. This allowed us to account for repeated measures within children across time and adjust for the baseline covariates of BMI, race, and calcium intake and for the time-varying covariates of 25(OH)D and season. In the models, season was based on the date of visit and categorized as fall/winter/early- to midspring (October through May), and late spring/summer (June through September). We assessed cross-sectional between-child associations using baseline measures and within-child associations using changes from baseline.
Results Of 304 eligible children, 157 were enrolled (Figure 1), 141 in Pittsburgh and 16 in Kittanning. Eighty-four of the children were black and 73 were white. The vitamin D3-supplemented and placebo groups were similar at baseline (Table 1). Baseline characteristics between vitamin D and placebo groups, when examined within each race group, were also balanced except for hand melanin index in black children (Table 1). However, hand melanin index measurements were missing in 12 black children. Effects of Vitamin D3 supplementation Effects on mean 25(OH)D concentrations Baseline 25(OH)D concentrations were not different between the vitamin D3-supplemented and placebo groups, and both were ⬍20 ng/mL (Table 2). At the 2- and
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 August 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.
3186
Rajakumar et al
Table 1.
Vitamin D Supplementation in Children
J Clin Endocrinol Metab, August 2015, 100(8):3183–3192
Baseline Characteristics of Enrolled Children All Children
n Demographic Male Black Ethnicity Hispanic Non-Hispanic Not reported Age, y Weight, kg Height, cm BMI, kg/m2 Weight classification Normal weight (BMI ⬍85th %tile) Overweight (BMI 85th to ⬍95th %tiles) Obese (BMI ⱖ95th %tile) Tanner stage I II III IV V Skin type I (easy burn, no tan) II (easy burn, slight tan) III (burn, then tan) IV (no burn, good tan) V (never burn, marked tan) Melanin indexa Forehead Hand Underarm Summertime sunlight exposure Duration ⬎ 2 h Sunscreen use, yes Sunscreen use frequency Often Sometimes Seldom Vacation travel to sunny location, yes Laboratory data Calcium, mg/dL Phosphorus, mg/dL Albumin, g/dL 25(OH)D, ng/mL PTH, pg/mL OC, ng/mL CTx, ng/mL Dietary intake Calcium, mg/d Vitamin D, IU/d Vitamin D-deficient 25(OH)D ⬍ 20 ng/mL
Black Children
Vitamin D Group
Placebo Group
78
79
33 (42.3) 42 (53.8)
45 (57.0) 42 (53.2)
1 (1.3) 68 (87.2) 9 (11.5) 11.2 ⫾ 1.9 48.4 ⫾ 18.0 148.1 ⫾ 13.4 21.6 ⫾ 5.5
White Children
Vitamin D Group
Placebo Group
Vitamin D Group
Placebo Group
42
42
36
37
.07 .93
18 (42.9)
23 (54.8)
.38
15 (41.7)
22 (59.5)
1 (1.3) 63 (79.7) 15 (19.0) 11.4 ⫾ 2.0 49.2 ⫾ 19.6 149.3 ⫾ 13.1 21.6 ⫾ 6.2
.42
1 (2.4) 33 (78.6) 8 (19.1) 11.1 ⫾ 1.9 49.7 ⫾ 18.2 147.8 ⫾ 12.4 22.3 ⫾ 6.1
1 (2.4) 30 (71.4) 11 (26.2) 11.8 ⫾ 2.0 54.8 ⫾ 23.3 153.2 ⫾ 13.8 22.8 ⫾ 7.5
.80
0 35 (97.2) 1 (2.8) 11.3 ⫾ 1.9 46.9 ⫾ 18.0 148.5 ⫾ 14.6 20.7 ⫾ 4.7
33 (89.2) 4 (10.8) 10.9 ⫾ 1.8 42.8 ⫾ 11.7 144.9 ⫾ 10.8 20.1 ⫾ 3.8
44 (56.4) 14 (17.9) 20 (25.6)
49 (62.0) 11 (13.9) 19 (24.1)
.72
20 (47.6) 8 (19.1) 14 (33.3)
25 (59.5) 5 (11.9) 12 (28.6)
.48
24 (66.7) 6 (16.7) 6 (16.7)
24 (64.9) 6 (16.2) 7 (18.9)
1.00
30 (38.5) 20 (25.6) 12 (15.4) 8 (10.3) 8 (10.3)
31 (39.2) 19 (24.1) 14 (17.7) 8 (10.1) 7 (8.9)
.99
14 (33.3) 12 (28.6) 6 (14.3) 6 (14.3) 4 (9.5)
13 (31.0) 6 (14.3) 10 (23.8) 6 (14.3) 7 (16.7)
.43
16 (44.4) 8 (22.2) 6 (16.7) 2 (5.6) 4 (11.1)
18 (48.7) 13 (35.1) 4 (10.8) 2 (5.4) 0 (0)
.25
4 (5.3) 14 (18.4) 15 (19.7) 36 (47.4) 7 (9.2)
6 (7.7) 13 (16.7) 16 (20.5) 27 (34.6) 16 (20.5)
.26
0 (0) 0 (0) 4 (9.8) 30 (73.2) 7 (17.1)
0 (0) 1 (2.4) 4 (9.8) 20 (48.8) 16 (39.0)
.06
4 (11.4) 14 (40.0) 11 (31.4) 6 (17.1) 0 (0)
6 (16.2) 12 (32.4) 12 (32.4) 7 (18.9) 0 (0)
.91
51.9 ⫾ 18.2 54.2 ⫾ 18.1 51.2 ⫾ 16.8
54.6 ⫾ 19.8 57.5 ⫾ 20.0 53.1 ⫾ 18.5
.40 .28 .52
66.9 ⫾ 13.5 69.0 ⫾ 13.5 64.6 ⫾ 13.7
71.9 ⫾ 11.4 75.6 ⫾ 10.7 69.2 ⫾ 12.2
.09 .02 .14
37.0 ⫾ 5.4 39.3 ⫾ 5.3 37.8 ⫾ 4.1
37.3 ⫾ 7.1 39.6 ⫾ 5.8 37.5 ⫾ 5.6
.83 .81 .75
63 (85.1) 28 (37.8)
60 (78.9) 36 (46.8)
.32 .27
33 (84.6) 6 (15.4)
35 (87.5) 13 (31.7)
.76 .12
30 (85.7) 22 (62.9)
25 (69.4) 23 (63.9)
.16 1.00
13 (54.2) 11 (45.8) 0 (0.0) 14 (18.9)
22 (68.8) 8 (25.0) 2 (6.3) 22 (28.6)
.32
4 (40) 4 (40) 2 (20) 9 (22.0)
.44
.12
12 (63.2) 7 (36.8) 0 (0) 11 (31.4)
18 (81.8) 4 (18.2) 0 (0) 13 (36.1)
.29
.16
1 (20) 4 (80) 0 (0) 3 (7.7)
.80
9.7 ⫾ 0.4 5.0 ⫾ 0.5 4.2 ⫾ 0.3 19.8 ⫾ 7.6 36.9 ⫾ 19.6 94.2 ⫾ 44.9 1.7 ⫾ 0.9
9.6 ⫾ 0.4 5.1 ⫾ 0.6 4.2 ⫾ 0.3 18.8 ⫾ 6.9 37.6 ⫾ 19.0 104.5 ⫾ 48.4 1.8 ⫾ 0.9
.43 .14 .64 .38 .83 .18 .82
9.8 ⫾ 0.38 4.9 ⫾ 0.49 4.3 ⫾ 0.33 16.6 ⫾ 7.4 43.4 ⫾ 20.8 88.0 ⫾ 41.9 2.1 ⫾ 1.0
9.7 ⫾ 0.39 5.1 ⫾ 0.58 4.2 ⫾ 0.34 16.3 ⫾ 6.7 42.3 ⫾ 21.0 93.4 ⫾ 54.5 2.0 ⫾ 1.0
.21 .31 .48 .81 .80 .62 .74
9.5 ⫾ 0.35 5.0 ⫾ 0.42 4.2 ⫾ 0.22 23.5 ⫾ 6.0 29.4 ⫾ 15.1 101.3 ⫾ 47.6 1.3 ⫾ 0.6
9.5 ⫾ 0.40 5.1 ⫾ 0.58 4.2 ⫾ 0.23 21.6 ⫾ 6.1 32.3 ⫾ 15.2 116.4 ⫾ 38.0 1.4 ⫾ 0.7
.83 .28 .79 .19 .41 .14 .30
1164.3 ⫾ 602.2 253.6 ⫾ 141.0
1200.6 ⫾ 599.3 283.9 ⫾ 175.0
.71 .23
1157.0 ⫾ 600.9 239.4 ⫾ 132.1
1178.7 ⫾ 695.5 262.2 ⫾ 166.9
.88 .49
1172.8 ⫾ 612.1 270.2 ⫾ 150.9
1226.1 ⫾ 471.5 309.3 ⫾ 182.9
.68 .33
42 (54)
45 (57)
.69
30 (71)
29 (69)
.81
12 (33)
16 (43)
.38
P Value
.48 .80 .59 .97
P Value
.08 .27 .06 .75
P Value
.16
.36
.35 .25 .22 .56
Data are expressed as number (percentage) or mean ⫾ SD, unless stated otherwise. a Melanin index from forehead and hand was not obtained in 11 black children (vitamin D group, n ⫽ 6; placebo group, n ⫽ 5), and melanin index from underarm was not obtained in 12 black children (vitamin D group, n ⫽ 6; placebo group, n ⫽ 6).
6-month follow-up visits, concentrations were higher in supplemented children, and in black supplemented children considered as a separate subgroup. The increase in 25(OH)D concentration in supplemented children was greater in black than in white children at 2 months but not at 6 months. Mean 25(OH)D concentrations were lower generally in black children than in white children. Analysis of covariance parameter estimates for the 25(OH)D differences at 2 and 6 months are shown in Supplemental Table 1.
Children with baseline 25(OH)D concentrations
ARTICLE
Effect of Vitamin D3 Supplementation in Black and in White Children: A Randomized, Placebo-Controlled Trial Kumaravel Rajakumar, Charity G. Moore, Jonathan Yabes, Flora Olabopo, Mary Ann Haralam, Diane Comer, Jaimee Bogusz, Anita Nucci, Susan Sereika, Jacqueline Dunbar-Jacob, Michael F. Holick, and Susan L. Greenspan Department of Pediatrics (K.R., F.O., M.A.H.), and Center for Research on Health Care (C.G.M., J.Y., D.C.), University of Pittsburgh, Pittsburgh, Pennsylvania 15213; Department of Medicine (J.B., M.F.H.), Boston University School of Medicine, Boston, Massachusetts, 02118; Department of Nutrition (A.N.), Georgia State University, Atlanta, Georgia 30302; University of Pittsburgh School of Nursing (S.S., J.D.-J.), Pittsburgh, Pennsylvania 15213; and Department of Medicine (C.G.M., J.Y., D.C., S.L.G.), University of Pittsburgh, Pittsburgh 15213
Context: Dosages of vitamin D necessary to prevent or treat vitamin D deficiency in children remain to be clarified. Objective: To determine the effects of vitamin D3 1000 IU/d on serum 25-hydroxyvitamin D [25(OH)D], PTH, and markers of bone turnover (osteocalcin and collagen type 1 cross-linked Ctelopeptide) in black children and white children, and to explore whether there is a threshold level of 25(OH)D associated with maximal suppression of serum PTH concentration. Design: Healthy 8- to 14-year-old Pittsburgh-area black (n ⫽ 84) and white (n ⫽ 73) children not receiving vitamin supplements, enrolled from October through March from 2008 through 2011, were randomized to vitamin D3 1000 IU or placebo daily for 6 months. Results: The mean baseline concentration of 25(OH)D was ⬍20 ng/mL in both the vitamin D-supplemented group and the placebo group (19.8 ⫾ 7.6 and 18.8 ⫾ 6.9 ng/mL, respectively). The mean concentration was higher in the supplemented group than in the placebo group at 2 months (26.4 ⫾ 8.1 vs 18.9 ⫾ 8.1 ng/mL; P ⬍ .0001) and also at 6 months (26.7 ⫾ 7.6 vs 22.4 ⫾ 7.3; P ⫽ .003), after adjusting for baseline 25(OH)D, race, gender, pubertal status, dietary vitamin D intake, body mass index, and sunlight exposure. Increases were only significant in black children, when examined by race. The association between 25(OH)D and PTH concentrations was inverse and linear, without evidence of a plateau. Overall, vitamin D supplementation had no effect on PTH and bone turnover. Conclusions: Vitamin D3 supplementation with 1000 IU/d in children with mean baseline 25(OH)D concentration ⬍20 ng/mL effectively raised their mean 25(OH)D concentration to ⱖ20 ng/mL but failed to reach 30 ng/mL. Vitamin D supplementation had no effect on PTH concentrations. (J Clin Endocrinol Metab 100: 3183–3192, 2015)
aintaining adequate vitamin D status is essential for calcium homeostasis and skeletal health. However, hypovitaminosis D is common in healthy children living in the northeastern United States, and its prevalence
M
and severity are greater in black than in white children (1). Circulating concentration of 25-hydroxyvitamin D [25(OH)D] is the recognized biomarker of vitamin D status. Definition of a 25(OH)D cutoff level for optimal skel-
ISSN Print 0021-972X ISSN Online 1945-7197 Printed in USA Copyright © 2015 by the Endocrine Society Received March 11, 2015. Accepted June 16, 2015. First Published Online June 19, 2015
Abbreviations: BMI, body mass index; CTx, collagen type 1 cross-linked C-telopeptide; CV, coefficient of variation; DBP, vitamin D-binding-protein; OC, osteocalcin; 25(OH)D, 25hydroxyvitamin D; ⌬, change.
doi: 10.1210/jc.2015-1643
J Clin Endocrinol Metab, August 2015, 100(8):3183–3192
press.endocrine.org/journal/jcem
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 August 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.
3183
3184
Rajakumar et al
Vitamin D Supplementation in Children
etal health lacks consensus. The Institute of Medicine (IOM) recommends concentrations of 25(OH)D ⱖ20 ng/mL as optimal for skeletal health and defines vitamin D deficiency as a concentration ⬍12 ng/mL and vitamin D insufficiency as concentrations 12 to 20 ng/mL (2). The IOM estimates that concentrations of ⱖ16 ng/mL and ⱖ20 ng/mL would be adequate for ensuring the skeletal health needs of 50 and 97.5%, respectively, of US children. The Endocrine Society and other experts in the field have suggested a target level ⱖ30 ng/mL for optimal skeletal health (3, 4). Concentrations of 25(OH)D associated with maximal suppression of serum PTH concentration, indicated by a plateauing of the PTH concentration, have been used for defining vitamin D sufficiency in adults. Such data in children are limited and varied. Hill et al (5) found no inflection point in the inverse association between 25(OH)D and PTH in a cross-sectional study of 7- to 18-year-old children. Although Maguire et al (6) demonstrated plateauing of PTH at a 25(OH)D concentration of 42.8 ng/mL in a cross-sectional study of 1- to 6-year-old children, vitamin D supplementation (0, 400, 1000, 2000, and 4000 IU/d for 12 wk) in 9- to 13-year-old children had no effect on PTH concentrations despite dose-dependent increases in 25(OH)D concentrations (7). Confounding effects of sun exposure and other determinants of vitamin D photoproduction pose challenges for estimating the amount of dietary vitamin D needed to achieve and maintain a defined 25(OH)D level (2). Those considerations notwithstanding, the IOM’s dietary reference intakes for vitamin D were calculated without regard for variations in skin color, race, or sunlight exposure. Delineating the racial differences in response to vitamin D supplementation and exploring levels of 25(OH)D associated with vitamin D sufficiency is relevant for formulating dietary guidelines. In addition, data regarding the effect of vitamin D supplementation on bone turnover are limited. Therefore, we initiated a pharmacological challenge with 1000 IU of vitamin D3 daily (five times the prevailing adequate intake for vitamin D) (8) in black children and white children to examine the responsive changes in 25(OH)D, PTH, and bone turnover. The objective of this study was to determine the effects of supplementation with 1000 IU of vitamin D3 on serum concentrations of 25(OH)D, PTH, and markers of bone turnover in black children and white children. We utilized the longitudinal design of our study to determine whether there was a threshold level of 25(OH)D associated with maximal suppression of serum PTH concentration.
J Clin Endocrinol Metab, August 2015, 100(8):3183–3192
Subjects and Methods Study design and participants We enrolled healthy 8- to 14-year-old children from Pittsburgh and Kittanning, Pennsylvania (latitude/longitude: 40.4° N/80°W and 40.8° N/79.5°W, respectively) in a randomized, double-blind, placebo-controlled trial of vitamin D3 1000 IU daily (Clinicaltrials.gov identifier: NCT00732758). The children were enrolled October through March of 2008 through 2011. Children receiving vitamin preparations underwent a 1-month washout before enrollment. Subjects were recruited from the Primary Care Center of the Children’s Hospital of Pittsburgh of the University of Pittsburgh Medical Center (UPMC) and the Children’s Community Pediatrics-Armstrong practice in Kittanning, and through advertisements posted in the offices of the Children’s Hospital-affiliated Pediatric PittNet, a practice-based pediatric research network. Study procedures subsequent to enrollment were conducted either at the UPMC Montefiore Clinical and Translational Research Center or at the Children’s Community Pediatrics-Armstrong practice. The study was approved by the University of Pittsburgh Institutional Review Board. Signed informed parental consent and subjects’ assent were obtained before research participation. The race of subjects was specified by their parents.
Randomization and intervention Randomization was stratified by race using a 1:1 allocation ratio and a permuted block scheme with block size of 4. Children received either vitamin D3 1000 IU or placebo in a single tablet once daily. The allocation scheme was generated by a study statistician using R version 2.7.0. The vitamin D3 and placebo tablets were manufactured by Douglas Laboratories, were similar in color, and were dispensed in identical containers labeled either A or B. A sealed envelope system was used for the assignments. Children, parents, and the investigative team were blinded to the treatment assignment until the study was completed. Study medications were made in two batches. The average vitamin D3 content of the vitamin D3 tablet, measured as previously described (9), in the first batch around the midpoint of its shelf-life was approximately 1129 IU; in the second batch at the end of the trial, it was approximately 1140 IU.
Compliance Compliance was assessed by pill count at the 2- and 6-month follow-up visits and was validated in a subset of 90 subjects by an electronic medication event monitoring system (MEMs 6 Track Cap; AARDEX).
Study measurements Anthropometry, sun exposure, skin color, dietary intake, and pubertal status We measured height and weight and calculated body mass index (BMI) at study entry and at the 6-month follow-up visit. At study entry and exit, we assessed: summertime sunlight exposure characteristics; melanin index from the forehead, back of the hand, and inner arm using a handheld dermatospectrophotometer (DSM II Colormeter; Cortex Technology); and dietary intake of vitamin D and calcium using a validated Youth and Adolescent Food Frequency Questionnaire (10, 11). We estimated
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 August 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2015-1643
press.endocrine.org/journal/jcem
Tanner stage by physical examination (12, 13) and ascertained the parent-reported Fitzpatrick sunreactive skin type (14 –16) at study entry.
3185
Study Flow Diagram
Assessed for eligibility (n=355)
Ineligible (n=51)
Biochemical assessments We collected blood samples by venipuncture in a nonfasting state throughout the day. We measured serum calcium, phosphorus, albumin, 25(OH)D, PTH, osteocalcin (OC; marker of bone formation), and collagen type 1 cross-linked C-telopeptide (CTx; a marker of bone resorption) concentrations at baseline and at 2- and 6-month follow-up visits. Calcium, phosphorus, and albumin concentrations were measured at the UPMC Clinical Chemistry laboratory. Serum 25(OH)D, PTH, OC, and CTx concentrations were measured at the Vitamin D, Skin, and Bone Research Laboratory at Boston University Medical Center. Serum 25(OH)D concentrations were measured using liquid chromatography-tandem mass spectrometry (17). The intra-assay and interassay coefficients of variation (CVs) were 6% and 10%, respectively. The 25(OH)D assay is Center for Disease Control certified, and National Institutes of Standards and Technology standards were used for confirmation of the standard curves. Serum 3-epi-25(OH)D is excluded in the reported levels. Serum PTH concentrations were assayed using a Human Bioactive PTH 1– 84 Elisa kit (Immutopics, Inc); intra-assay and interassay CVs were 7 and 9%, respectively. Serum OC concentrations were measured by Micro Vue OC enzyme immunoassay kit (Quidel); intra-assay and interassay CVs were 5 and 10%, respectively. Serum CTx concentrations were measured by Serum CrossLaps ELISA kit (Immunodiagnostic Systems); intra-assay and interassay CVs were 1.8 –3% and 8.0 –10.9%, respectively.
Statistical analysis We based parameter estimates for sample-size analyses on our earlier study of changes in PTH concentration in response to vitamin D3 supplementation (18). Assuming that the concentration of 25(OH)D would reach ⬎20 ng/mL in the supplemented group and would remain ⱕ20 ng/mL in the placebo group and with two-sided ␣ ⫽ 0.05, we estimated that 160 children would be needed to detect effect sizes in serum 25(OH)D and PTH concentrations of approximately 0.5, with power of 81 to 88%. Assuming that 50% of the children were black and 50% white, we had 79% power to detect a correlation of at least ⫺0.3 between PTH and 25(OH)D concentrations within each race group (␣ ⫽ 0.05). We anticipated an attrition rate of 5% and planned to enroll a total of 168 children. We used intention-to-treat analyses when testing the effects of vitamin D supplementation relative to placebo. We tested for treatment group or racial differences in categorical measures using 2 or Fisher’s exact tests and in continuous measures using t tests. Concentrations of 25(OH)D, the primary study outcome, did not require transformation. We compared mean 25(OH)D concentrations using analysis of covariance, controlling for baseline 25(OH)D, race, gender, pubertal status, BMI, dietary intake of vitamin D, and sunlight exposure. To explore threshold effects of 25(OH)D on PTH concentrations, we conducted analyses similar to those described by Hill et al (5) and Maguire et al (6) using linear and cubic spline regressions. Scatterplots were constructed superimposed with both fitted lines and Lowess (locally weighted scatterplot smoothing) curves to visually examine nonlinearity. We used mixed-effects modeling to examine the associations between
Eligible (n=304)
Did not agree to parcipate (n=147)
Consent form signed (n=157)
Randomized (n=157)
Vitamin D Group (n=78) Black, 42; White, 36
Lost to follow-up (n=8)
Completed (n=70)
Placebo Group (n=79) Black, 42; White, 37
Lost to follow-up (n=16)
Completed (n=63)
Figure 1. Enrollment, randomization, and follow-up of study participants
PTH and 25(OH)D concentrations. This allowed us to account for repeated measures within children across time and adjust for the baseline covariates of BMI, race, and calcium intake and for the time-varying covariates of 25(OH)D and season. In the models, season was based on the date of visit and categorized as fall/winter/early- to midspring (October through May), and late spring/summer (June through September). We assessed cross-sectional between-child associations using baseline measures and within-child associations using changes from baseline.
Results Of 304 eligible children, 157 were enrolled (Figure 1), 141 in Pittsburgh and 16 in Kittanning. Eighty-four of the children were black and 73 were white. The vitamin D3-supplemented and placebo groups were similar at baseline (Table 1). Baseline characteristics between vitamin D and placebo groups, when examined within each race group, were also balanced except for hand melanin index in black children (Table 1). However, hand melanin index measurements were missing in 12 black children. Effects of Vitamin D3 supplementation Effects on mean 25(OH)D concentrations Baseline 25(OH)D concentrations were not different between the vitamin D3-supplemented and placebo groups, and both were ⬍20 ng/mL (Table 2). At the 2- and
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 August 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.
3186
Rajakumar et al
Table 1.
Vitamin D Supplementation in Children
J Clin Endocrinol Metab, August 2015, 100(8):3183–3192
Baseline Characteristics of Enrolled Children All Children
n Demographic Male Black Ethnicity Hispanic Non-Hispanic Not reported Age, y Weight, kg Height, cm BMI, kg/m2 Weight classification Normal weight (BMI ⬍85th %tile) Overweight (BMI 85th to ⬍95th %tiles) Obese (BMI ⱖ95th %tile) Tanner stage I II III IV V Skin type I (easy burn, no tan) II (easy burn, slight tan) III (burn, then tan) IV (no burn, good tan) V (never burn, marked tan) Melanin indexa Forehead Hand Underarm Summertime sunlight exposure Duration ⬎ 2 h Sunscreen use, yes Sunscreen use frequency Often Sometimes Seldom Vacation travel to sunny location, yes Laboratory data Calcium, mg/dL Phosphorus, mg/dL Albumin, g/dL 25(OH)D, ng/mL PTH, pg/mL OC, ng/mL CTx, ng/mL Dietary intake Calcium, mg/d Vitamin D, IU/d Vitamin D-deficient 25(OH)D ⬍ 20 ng/mL
Black Children
Vitamin D Group
Placebo Group
78
79
33 (42.3) 42 (53.8)
45 (57.0) 42 (53.2)
1 (1.3) 68 (87.2) 9 (11.5) 11.2 ⫾ 1.9 48.4 ⫾ 18.0 148.1 ⫾ 13.4 21.6 ⫾ 5.5
White Children
Vitamin D Group
Placebo Group
Vitamin D Group
Placebo Group
42
42
36
37
.07 .93
18 (42.9)
23 (54.8)
.38
15 (41.7)
22 (59.5)
1 (1.3) 63 (79.7) 15 (19.0) 11.4 ⫾ 2.0 49.2 ⫾ 19.6 149.3 ⫾ 13.1 21.6 ⫾ 6.2
.42
1 (2.4) 33 (78.6) 8 (19.1) 11.1 ⫾ 1.9 49.7 ⫾ 18.2 147.8 ⫾ 12.4 22.3 ⫾ 6.1
1 (2.4) 30 (71.4) 11 (26.2) 11.8 ⫾ 2.0 54.8 ⫾ 23.3 153.2 ⫾ 13.8 22.8 ⫾ 7.5
.80
0 35 (97.2) 1 (2.8) 11.3 ⫾ 1.9 46.9 ⫾ 18.0 148.5 ⫾ 14.6 20.7 ⫾ 4.7
33 (89.2) 4 (10.8) 10.9 ⫾ 1.8 42.8 ⫾ 11.7 144.9 ⫾ 10.8 20.1 ⫾ 3.8
44 (56.4) 14 (17.9) 20 (25.6)
49 (62.0) 11 (13.9) 19 (24.1)
.72
20 (47.6) 8 (19.1) 14 (33.3)
25 (59.5) 5 (11.9) 12 (28.6)
.48
24 (66.7) 6 (16.7) 6 (16.7)
24 (64.9) 6 (16.2) 7 (18.9)
1.00
30 (38.5) 20 (25.6) 12 (15.4) 8 (10.3) 8 (10.3)
31 (39.2) 19 (24.1) 14 (17.7) 8 (10.1) 7 (8.9)
.99
14 (33.3) 12 (28.6) 6 (14.3) 6 (14.3) 4 (9.5)
13 (31.0) 6 (14.3) 10 (23.8) 6 (14.3) 7 (16.7)
.43
16 (44.4) 8 (22.2) 6 (16.7) 2 (5.6) 4 (11.1)
18 (48.7) 13 (35.1) 4 (10.8) 2 (5.4) 0 (0)
.25
4 (5.3) 14 (18.4) 15 (19.7) 36 (47.4) 7 (9.2)
6 (7.7) 13 (16.7) 16 (20.5) 27 (34.6) 16 (20.5)
.26
0 (0) 0 (0) 4 (9.8) 30 (73.2) 7 (17.1)
0 (0) 1 (2.4) 4 (9.8) 20 (48.8) 16 (39.0)
.06
4 (11.4) 14 (40.0) 11 (31.4) 6 (17.1) 0 (0)
6 (16.2) 12 (32.4) 12 (32.4) 7 (18.9) 0 (0)
.91
51.9 ⫾ 18.2 54.2 ⫾ 18.1 51.2 ⫾ 16.8
54.6 ⫾ 19.8 57.5 ⫾ 20.0 53.1 ⫾ 18.5
.40 .28 .52
66.9 ⫾ 13.5 69.0 ⫾ 13.5 64.6 ⫾ 13.7
71.9 ⫾ 11.4 75.6 ⫾ 10.7 69.2 ⫾ 12.2
.09 .02 .14
37.0 ⫾ 5.4 39.3 ⫾ 5.3 37.8 ⫾ 4.1
37.3 ⫾ 7.1 39.6 ⫾ 5.8 37.5 ⫾ 5.6
.83 .81 .75
63 (85.1) 28 (37.8)
60 (78.9) 36 (46.8)
.32 .27
33 (84.6) 6 (15.4)
35 (87.5) 13 (31.7)
.76 .12
30 (85.7) 22 (62.9)
25 (69.4) 23 (63.9)
.16 1.00
13 (54.2) 11 (45.8) 0 (0.0) 14 (18.9)
22 (68.8) 8 (25.0) 2 (6.3) 22 (28.6)
.32
4 (40) 4 (40) 2 (20) 9 (22.0)
.44
.12
12 (63.2) 7 (36.8) 0 (0) 11 (31.4)
18 (81.8) 4 (18.2) 0 (0) 13 (36.1)
.29
.16
1 (20) 4 (80) 0 (0) 3 (7.7)
.80
9.7 ⫾ 0.4 5.0 ⫾ 0.5 4.2 ⫾ 0.3 19.8 ⫾ 7.6 36.9 ⫾ 19.6 94.2 ⫾ 44.9 1.7 ⫾ 0.9
9.6 ⫾ 0.4 5.1 ⫾ 0.6 4.2 ⫾ 0.3 18.8 ⫾ 6.9 37.6 ⫾ 19.0 104.5 ⫾ 48.4 1.8 ⫾ 0.9
.43 .14 .64 .38 .83 .18 .82
9.8 ⫾ 0.38 4.9 ⫾ 0.49 4.3 ⫾ 0.33 16.6 ⫾ 7.4 43.4 ⫾ 20.8 88.0 ⫾ 41.9 2.1 ⫾ 1.0
9.7 ⫾ 0.39 5.1 ⫾ 0.58 4.2 ⫾ 0.34 16.3 ⫾ 6.7 42.3 ⫾ 21.0 93.4 ⫾ 54.5 2.0 ⫾ 1.0
.21 .31 .48 .81 .80 .62 .74
9.5 ⫾ 0.35 5.0 ⫾ 0.42 4.2 ⫾ 0.22 23.5 ⫾ 6.0 29.4 ⫾ 15.1 101.3 ⫾ 47.6 1.3 ⫾ 0.6
9.5 ⫾ 0.40 5.1 ⫾ 0.58 4.2 ⫾ 0.23 21.6 ⫾ 6.1 32.3 ⫾ 15.2 116.4 ⫾ 38.0 1.4 ⫾ 0.7
.83 .28 .79 .19 .41 .14 .30
1164.3 ⫾ 602.2 253.6 ⫾ 141.0
1200.6 ⫾ 599.3 283.9 ⫾ 175.0
.71 .23
1157.0 ⫾ 600.9 239.4 ⫾ 132.1
1178.7 ⫾ 695.5 262.2 ⫾ 166.9
.88 .49
1172.8 ⫾ 612.1 270.2 ⫾ 150.9
1226.1 ⫾ 471.5 309.3 ⫾ 182.9
.68 .33
42 (54)
45 (57)
.69
30 (71)
29 (69)
.81
12 (33)
16 (43)
.38
P Value
.48 .80 .59 .97
P Value
.08 .27 .06 .75
P Value
.16
.36
.35 .25 .22 .56
Data are expressed as number (percentage) or mean ⫾ SD, unless stated otherwise. a Melanin index from forehead and hand was not obtained in 11 black children (vitamin D group, n ⫽ 6; placebo group, n ⫽ 5), and melanin index from underarm was not obtained in 12 black children (vitamin D group, n ⫽ 6; placebo group, n ⫽ 6).
6-month follow-up visits, concentrations were higher in supplemented children, and in black supplemented children considered as a separate subgroup. The increase in 25(OH)D concentration in supplemented children was greater in black than in white children at 2 months but not at 6 months. Mean 25(OH)D concentrations were lower generally in black children than in white children. Analysis of covariance parameter estimates for the 25(OH)D differences at 2 and 6 months are shown in Supplemental Table 1.
Children with baseline 25(OH)D concentrations
